14 BULLETIN 875, U. S. DEPARTMENT OF AGRICULTURE. 
chine investment per acre too high. On the other hand, it is un- 
doubtedly the wisest policy to utilize methods which will give the 
greatest possible degree of control during the first few years when it 
is necessary for everyone to feel his way, more or less, in arranging 
the poisoning schedule. Consequently, it is urged that poisoning be 
attempted only under such conditions as will justify a sufficient 
machinery outlay to permit poisoning at about a four-day time 
interval. 
NUMBER OF APPLICATIONS. 
The question as to the number of applications is closely correlated 
with that of the time interval and is one which every man must de- 
cide for himself by experience, as the number will vary widely under 
different conditions. In the large Delta plantations, where the ma- 
jority of the work has been conducted in the past, it has been found 
that weevil control throughout the season in the cuts most heavily in- 
tested and requiring the earliest treatment in the spring has necessi- 
tated from four to six applications with a time interval of one week. 
Other cuts more distant from hibernation quarters and thus more 
lightly infested required three, two, one, or no applications. In two 
years’ work on thousands of acres handled on this basis, the average 
number of applications for all cotton acreage involved has been 
something less than two. Where infestation is more uniformly dis- 
tributed at the outset and general poisoning is necessary, it has been 
found advisable to figure on an average of about four applications, 
and in case of an excessively rainy season, five or six applications. 
These figures are based on a one-week time interval, as this interval 
has been the one adopted in the majority of work so far. With a 
chorter time interval, however, control can be secured with about 
three applications. This office is now planning to open a number of 
cmall experiment stations in nearly all representative districts of 
the cotton belt. By conducting control tests at all of these at the 
same time it will soon be possible to designate much more definite 
rules of procedure under the varying conditions found in these 
different districts. 
TIME TO STCP POISONING. 
The time to step poisoning has been more or less covered by the 
discussion of the number of applications but it is probably well 
to outline briefly the conditions governing this. The idea is to 
maintain weevil control below the point of loss to the crop long 
enough for the plants to set as many bolls as they can mature and 
also to protect these beyond the danger of weevil puncturing. 
Furthermore, there is nearly always a time when the plants cease 
to retain any more fruit. In many cases they not only discontinue 
squaring and blooming but shed the young bolls as fast as they are 
